How to use Python library in Julia?

Specific Domains Machine Learning
, ,
1

I was wondering how can I use this package https://github.com/DeepGraphLearning/ConfGF in Julia by using PyCall.jl or PythonCall.jl?

this is how to use this package in python. I want to do it in Julia, is it feasible?

# Create conda environment
conda create -n confgf python=3.7

# Activate the environment
conda activate confgf

# Install packages
conda install -y -c pytorch pytorch=1.7.0 torchvision torchaudio cudatoolkit=10.2
conda install -y -c rdkit rdkit==2020.03.2.0
conda install -y scikit-learn pandas decorator ipython networkx tqdm matplotlib
conda install -y -c conda-forge easydict
pip install pyyaml

# Install PyTorch Geometric
pip install torch-scatter -f https://pytorch-geometric.com/whl/torch-1.7.0+cu102.html
pip install torch-sparse -f https://pytorch-geometric.com/whl/torch-1.7.0+cu102.html
pip install torch-cluster -f https://pytorch-geometric.com/whl/torch-1.7.0+cu102.html
pip install torch-spline-conv -f https://pytorch-geometric.com/whl/torch-1.7.0+cu102.html
pip install torch-geometric==1.6.3

# Install Library
git clone https://github.com/DeepGraphLearning/ConfGF.git
cd ConfGF
python setup.py install

it also needs command line to train the model.

Thank you in advance

2

Because of the heavy amount of dependencies, it sounds like pythoncall would be your best bet.

3

Thank you, from the tutorial of PythonCall.jl, it seems there is no material for this kind of application, It would be helpful if anyone can share some successful experience?

4

It depends what you’re trying to achieve, but if you already have all that stuff installed into a Conda environment then you can do

ENV["JULIA_PYTHONCALL_EXE"] = "python"
using PythonCall
@py import confgf
...

That first line forces PythonCall to use Python from your own Conda environment.

5

No I have not install all the things, I was thinking there is conda in Julia package, can I install all the things with that conda? and use this library entirely in Julia. I want to train the model with my data in Julia and produce some results in Julia, in other words, I want to use this library without leaving Julia. Thank you very much for your help!

python -u script/train.py --config_path ./config/qm9_default.yml
python -u script/train.py --config_path ./config/drugs_default.yml
python -u script/train.py --config_path ./config/iso17_default.yml

It also needs command line to train the model like that

If I install all these in a python environment, it need to create a virtual environment (“confgf”)

So we just import this library from this virtual environment in Julia with PythonCall?

How about the command line trainning part?

6

I installed the packages , and did as following:

ENV["JULIA_PYTHONCAL_EXE"] = "/usr/local/Caskroom/miniconda/base/envs/confgf/bin/python"
using PythonCall
@py import confgf

but it raised that confgf module not found error, after I tried torch, pandas or other packages installed there, same error,
only re, os this kind of module is ok.

the following code showed that confgf is installed.

Installed /usr/local/Caskroom/miniconda/base/envs/confgf/lib/python3.8/site-packages/confgf-0.1.0-py3.8.egg
Processing dependencies for confgf==0.1.0
Finished processing dependencies for confgf==0.1.0

What could be the problem? Thank you!

7

There are two Ls in PythonCall :slight_smile:

8

Interestingly, They gave the SAME result, after adding L
import os is ok but import confgf or torch or other packages same Module not found Error.

9

I found you have a great package CondaPkg.jl, do I need to create the environment and install all things using your CondaPkg.jl?

10

I reboot all of them and now it works, thank you very much!!!

BTW, can it use python command line?
like

python -u script/train.py --config_path ./config/qm9_default.yml
python -u script/train.py --config_path ./config/drugs_default.yml
python -u script/train.py --config_path ./config/iso17_default.yml